MIME-Version: 1.0
From: Mike Bushroe <mbushroe AT gmail DOT com>
Date: Sun, 16 Nov 2014 13:01:26 -0700
Message-ID: <CA+CyYa2Ps-Ehth-tYM15MudoPtcYkDGiyq5EJzPVVGt7vjsZEA@mail.gmail.com>
Subject: [geda-help] sch2pcb error log messages, possible extra info?
To: geda-help AT delorie DOT com
Content-Type: multipart/alternative; boundary=001a113d42dc8c619e0507ff5520
Reply-To: geda-help AT delorie DOT com

--001a113d42dc8c619e0507ff5520
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

gschem 1.8.2 (g875406c)
gsch2pcb 1.6
xgsch2pcb 0.1.3
pcb version 20140316 Compiled on Aug 29 2014 at 07:56:58
Ubuntu 14.10 (just up from 10.04 through 12.04)
Toshiba 5.7GiB Core 2 Duo 2.2GHzx2 64bit

I have bollixed up the schematics to the point where the error log coming
out of sch2pcb (message log window in pcb if it does not crash from the
more severe errors) seems to be thousands of lines long. After trying a
number of small changes to see if I am editing the same schematics it is
reading from, and making one or two lines actually disappear, and going to
running gnetlist -v manually I have begun to guess at the meaning of some
of the lines. the most common is "Can't find U12 pin 3 called for in
netlist.". When I open the net list in gedit, I can find device U12 pin 3
easily, and even see that it is in netlist Vcc, so I am confused when this
error shows up. I am now guessing that this means that U12 pin 3 is being
expected in more than one netlist, and its lack of presence in one list but
presence in another is why it is being called out. If the error message
could be expanded to indicate *where* it is not being found, such as in the
U1 pin 10 Vcc which has no other connections. If I could see in the regular
error log that it was looking for the Vcc net off of device U1 pin 10 that
it failed to find U12 pin 3, I might have been able to find out that my
hidden power connections had run a amok sooner that when I finally did. Of
course, that still leaves the matter of what to suggest to fix such errors,
but finding a universal solution to a syntax problem is even harder than
finding a universal description of the error in the first place! :)

  I also noticed with earlier naming convention problems that I would get
"can't find V pin s, pin number begins with letter, probably bad file
format". That turned out to be spaces in the pin label that caused the name
to shorten to just 'v', and the next usable character to be interpreted as
the pin number. Since I did not have a device "v" it took a while to track
down. If the internal communication from gnetlist to sch2pcb could add the
full text in the refdes attribute, or the schematic x,y location at the end
of each line so that when a defective device and pin number combination is
found that added reference data can be placed in the error log line to help
the user track down the miscreant symbol!

  I know that the normal response to 'why don't you make this change?' is
'why don't you make the changes yourself and upload them?'. I have never
tried to write code for an open source project before let alone up load
changes to a 'git' repository and do whatever happens to 'commit' the
changes. And right now I am *way* behind on getting what it little more
than a glorified copy of the device's sample circuit actually designed,
captured, *converted to pcb*, and laid out, so I want to get that done
first on my next re-rewrite of the schematic. And over the last year there
have been some times where looking through the details caused design
changes and outright electrical engineering errors to be caught so it has
not all  been bad. I but I am way overdue for my first piece of etched
copper! After that, I might actually think about making some minor code
changes in non-critical code sections and helping to flesh out the docs,
FAQ, and wiki and update the tutorials that we can.

Mike

--=20
"Creativity is intelligence having fun." =E2=80=94 Albert Einstein

--001a113d42dc8c619e0507ff5520
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div><div><div><div>gschem 1.8.2 (g875406c)<br>gsch2pcb 1.=
6<br>xgsch2pcb 0.1.3<br>pcb version 20140316 Compiled on Aug 29 2014 at 07:=
56:58<br>Ubuntu 14.10 (just up from 10.04 through 12.04)<br>Toshiba 5.7GiB =
Core 2 Duo 2.2GHzx2 64bit<br><br></div>I have bollixed up the schematics to=
 the point where the error log coming out of sch2pcb (message log window in=
 pcb if it does not crash from the more severe errors) seems to be thousand=
s of lines long. After trying a number of small changes to see if I am edit=
ing the same schematics it is reading from, and making one or two lines act=
ually disappear, and going to running gnetlist -v manually I have begun to =
guess at the meaning of some of the lines. the most common is &quot;Can&#39=
;t find U12 pin 3 called for in netlist.&quot;. When I open the net list in=
 gedit, I can find device U12 pin 3 easily, and even see that it is in netl=
ist Vcc, so I am confused when this error shows up. I am now guessing that =
this means that U12 pin 3 is being expected in more than one netlist, and i=
ts lack of presence in one list but presence in another is why it is being =
called out. If the error message could be expanded to indicate <i>where</i>=
 it is not being found, such as in the U1 pin 10 Vcc which has no other con=
nections. If I could see in the regular error log that it was looking for t=
he Vcc net off of device U1 pin 10 that it failed to find U12 pin 3, I migh=
t have been able to find out that my hidden power connections had run a amo=
k sooner that when I finally did. Of course, that still leaves the matter o=
f what to suggest to fix such errors, but finding a universal solution to a=
 syntax problem is even harder than finding a universal description of the =
error in the first place! :)<br><br></div>=C2=A0 I also noticed with earlie=
r naming convention problems that I would get &quot;can&#39;t find V pin s,=
 pin number begins with letter, probably bad file format&quot;. That turned=
 out to be spaces in the pin label that caused the name to shorten to just =
&#39;v&#39;, and the next usable character to be interpreted as the pin num=
ber. Since I did not have a device &quot;v&quot; it took a while to track d=
own. If the internal communication from gnetlist to sch2pcb could add the f=
ull text in the refdes attribute, or the schematic x,y location at the end =
of each line so that when a defective device and pin number combination is =
found that added reference data can be placed in the error log line to help=
 the user track down the miscreant symbol!<br><br></div>=C2=A0 I know that =
the normal response to &#39;why don&#39;t you make this change?&#39; is =
9;why don&#39;t you make the changes yourself and upload them?&#39;. I have=
 never tried to write code for an open source project before let alone up l=
oad changes to a &#39;git&#39; repository and do whatever happens to &#39;c=
ommit&#39; the changes. And right now I am <i>way</i> behind on getting wha=
t it little more than a glorified copy of the device&#39;s sample circuit a=
ctually designed, captured, <b>converted to pcb</b>, and laid out, so I wan=
t to get that done first on my next re-rewrite of the schematic. And over t=
he last year there have been some times where looking through the details c=
aused design changes and outright electrical engineering errors to be caugh=
t so it has not all=C2=A0 been bad. I but I am way overdue for my first pie=
ce of etched copper! After that, I might actually think about making some m=
inor code changes in non-critical code sections and helping to flesh out th=
e docs, FAQ, and wiki and update the tutorials that we can.<br><br></div>Mi=
ke<br clear=3D"all"><div><div><div><div><br>-- <br><div class=3D"gmail_sign=
ature"><div dir=3D"ltr"><span style=3D"font-family:arial,sans-serif;font-si=
ze:11pt">&quot;Creativity is intelligence having fun.&quot; =E2=80=94 Alber=
t Einstein</span></div></div>
</div></div></div></div></div>

--001a113d42dc8c619e0507ff5520--